Sustainable fisheries management requires the evaluation of the status of exploited populations and communities. Certain bycatch species such as rays are managed together which has led to overexploitation and conservations concerns for the most vulnerable ones. Traditional fisheries stock evaluation methods need species-specific input data while for ray species-specific landings have become available only recently in Europe. To overcome data limitations caused by grouped landings we developed a Bayesian multispecies biomass production model. In addition to the grouped landings the input data are shorter time series with species-specific information (landings and scientific biomass indices). Life history traits are used to define informative priors for intrinsic population growth rates. Applying the approach to the six main ray species managed together in the Bay of Biscay we identified long-term changes in community composition. Since the 1990s the small offshore cuckoo rays became increasingly dominant, while the contributions of the other five species declined, strongly for the largest species. In a prospective analysis, we found that all but the two largest species are expected to increase to biomass levels corresponding to maximum sustainable yield over the next decade under the current harvest rates while dedicated management will be needed to recover the strongly depleted largest species.